Re: [PATCH v6 1/6] i2c: Allow an ACPI driver to manage the device's power state during probe

From: Sakari Ailus <sakari.ailus@linux.intel.com>
To: Wolfram Sang <wsa@kernel.org>,
	linux-i2c@vger.kernel.org,
	"Rafael J. Wysocki" <rafael@kernel.org>,
	linux-acpi@vger.kernel.org, Bingbu Cao <bingbu.cao@intel.com>,
	linux-media@vger.kernel.org,
	Chiranjeevi Rapolu <chiranjeevi.rapolu@intel.com>,
	Hyungwoo Yang <hyungwoo.yang@intel.com>,
	Bartosz Golaszewski <bgolaszewski@baylibre.com>,
	Arnd Bergmann <arnd@arndb.de>,
	linux-kernel@vger.kernel.org,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	rajmohan.mani@intel.com, Tomasz Figa <tfiga@chromium.org>,
	"Qiu, Tian Shu" <tian.shu.qiu@intel.com>
Subject: Re: [PATCH v6 1/6] i2c: Allow an ACPI driver to manage the device's power state during probe
Date: Mon, 31 Aug 2020 12:54:15 +0300	[thread overview]
Message-ID: <20200831095415.GG31019@paasikivi.fi.intel.com> (raw)
In-Reply-To: <20200831093109.GA4632@kunai>

On Mon, Aug 31, 2020 at 11:31:09AM +0200, Wolfram Sang wrote:
> 
> > This patchset is really about changing the default of ACPI powering up I²C
> > devices. On OF the drivers are indeed responsible for that.
> 
> So, maybe it makes sense then to move from 'device_property_present()'
> to 'acpi_dev_get_property()' or something alike? To clearly tell this

I'll do that for v7 soon.

> binding is expected to be used with ACPI only. Then, we can skip this
> discussion now and postpone it to when someone wants to use it with DT.
> Which is hopefully never. Or does this approach have drawbacks?

The same issue in principle could be there on DT, too, as the cameras are
the same. There are a few sensor drivers supporting DT that currently don't
access the device in probe to avoid having to power it on. For cameras I
suppose that's just fine but I'd be hesitant changing the behaviour of e.g.
the at24 driver to support that use case without making it somehow
configurable.

> 
> > My original series had a field in struct device_driver for this purpose but
> > Greg K-H suggested moving it to I²C instead:
> > 
> > <URL:https://lore.kernel.org/linux-acpi/20190826084343.GA1095@kroah.com/>
> 
> Ok, we can still factor it out in the unlikely case it needs to be done
> again.
> 
> I still have the question via which tree this should go upstream?
> It is probably more I2C than ACPI?

I think so.

Rafael, would you be fine with this set being merged through the I²C tree?
There's a single patch adding an ACPI function there.

-- 
Sakari Ailus